← Reshaping Workflows with Dell Pro Precision and NVIDIA RTX PRO GPUs
Reshaping Workflows with Dell Pro Precision and NVIDIA RTX PRO GPUs
Live from GTC: Teaching A Robot to Dance with Mitch Chaiet
00:08:23
Founder/Senior Technologist at Experiential Technologies Mitch Chaiet joins Logan to share his story, how he programs his robot Sparky, and touch on the rapid speed at which robot tech is changing day-to-day.
Loading summary
Transcript12 lines
- [00:00]
A
Foreign.
- [00:05]
B
Welcome to reshaping workflows with Dell Pro Precision and Nvidia, where innovation meets real world impact in high performance computing.
- [00:20]
A
Hey, welcome. We are live at GTC 2026. We're actually outside in the GTC park. It's actually beautiful out here today. It's not super hot at this time of day. And guess what I got the man, the myth, the robot legend, the man himself, Mitch from xpt. Mitch, tell everyone what you do and why you love robots so much.
- [00:41]
C
Hello, everybody. My name is Mitch Chiat. I have a film degree and over the past eight months, somehow have found myself among the burgeoning humanoid robot industry. With my background in film and audio, I've found that the combination of microphones and cameras and LIDAR and mocap and VR and simulation and physical computing are sort of all rolled into one physical device, which is a robot. I grew up playing guitar and then started building my own guitars and then started making my own music on a computer and building my own instruments like MIDI controllers, which was a lot of fun. Which got me into things like Arduino, where I could start to code and build my own musical hardware and software. Fast forward five, 10 years. Was blessed with two Unitree G1 humanoid robots and a Dell T2 workstation, which has a fantastic Nvidia graphics card and suite of robotic simulation software, which has then allowed me to take motion capture data of a dancer and somebody doing karate. And I've just released a pack of G1 moves, which is 60 dance and karate moves trained from motion capture data all the way to a robot policy, which means the robot can actually do those moves. And so the beauty of that is you can film somebody or you can use an actual motion capture system. And that just gives you, like, data like you would use in a video game. The hard part is the robot has to perform that in real life, right? So much like a dancer learning a new choreography, you take that data, and then using some Nvidia tooling, you take that data and then you simulate the robot doing that dance move like 100,000 times. So cumulatively, you have like 700 years of GPU time of where a robot gets up, tries to do the dance move, and then falls over. And every one of those falling over starts to create a machine learning model called a policy. And that policy ends up becoming something that the robot can play back while it's balancing and then do that movement. So in the case of the dance moves we're seeing here at gtc, those are all policies trained from motion Capture data that then become a little file that you can load onto the robot and have it do that dance move in real life. So that's how I've taken my background in pure creative, gone all the way to playing with the best in the forefront of humanoid robotics software from my garage. So thanks to Dell and Nvidia for all the gear to make that happen.
- [03:36]
A
So Mitch obviously knows what he's doing most of the time. So, Mitch, I want to ask a question. Tell us a little bit about the robot badge you had to procure for the Unitree robot from the. What I like to dub the robot police.
- [03:52]
C
Correct. Well, first off, I'd like to thank everyone at Nvidia for all of the amazing robotic software, especially the folks running this amazing conference. GTC is my first year here, and I'm having a great time. For everyone who doesn't know, having 100 robots at a large conference requires some oversight. If you bring a robot that can do karate, we should tell the proper authorities. So for the first day and a half of gtc, we've been walking around here at the Dell booth performing amazing acts of dancing and karate at the forefront of humanoid robotics on the Unitree G1, which sparked lots of social media interest and attention from many of the amazing conference attendees and those organizing the conference. And we now have an approved robot badge for Sparky, the Unitree G1. It's an official conference attendee. Shout out to everyone on the Nvidia robotics team for helping us out there. Appreciate it.
- [04:54]
A
You're so great, Mitchell. All right. Last thing. You've done a great job. I mean, managing the robot. I think Trey's in charge now, and we can tell that it's not going as good. I will say this. Let me ask you a question. Every day you're texting me, hey, here's this new toolkit. Here's this new thing. In the terms of robotics versus AI, versus, you know, all the other things related to Nvidia, would you say robotics is evolving the fastest out of everything? Or, like, where would you put it? How do you keep up is the question.
- [05:25]
C
So I have no prior experience with proper robotics as of August, September 2025, which is kind of amazing for how quickly I've gotten up to speed with it. There have been times I've set out to do something like teleoperation, which is where you put on a VR headset. It uses body tracking to track your hands and then control the hands of the robot. There have been times where I set out on, like, a Monday to try and figure out how to do something. Then Nvidia will release something on Wednesday of that same week that didn't exist on Monday when I started it out. By Friday, there will be enough comments and feedback to where they'll squash a couple bugs. And then by Monday the next week, after a weekend of hacking, I've been able to get it running. So the software is evolving at such a rapid pace that I found it actually very interesting. Where I grew up on Internet forums learning how to hack, and I've always been of the mindset that you need to contribute back to the forum so other people can learn. And through things like Twitter, the Roboverse, Discord, I've met some amazing people. So shout out to Logan Olson, Harrison Kinsley, a lot of the great people posting amazing content with the Unitree G1. I'd like to thank Kevin Zaka, who's created an amazing package called MJLab, which is open source and that's how I've trained robot policies. And all of this great information and software is available free of charge. You just got to have the computer and the robot to do it. Even if you don't, you can still download it and run it for free and still learn in simulation. And I think that's the beauty of the community right now is if you post something cool, people will see it. If you post something interesting, you'll connect with the three people that actually still have robots that can utilize it, and then they will. So the software changes at such a fast pace that I have to refresh GitHub every morning. And that's what I love about robotics right now in 2026.
- [07:31]
A
Oh, Mitch. Well, with that, Mitch, appreciate you, man. You're awesome. And if you're at GTC listening to this, please stop by the Dell activation booth. You'll get to see see Sparky. Spark Sparky. The Dell AI factory robot.
- [07:44]
C
Yes.
- [07:45]
A
In action. Maybe. Maybe doing judo kicks. Maybe Chuck. Nor in your face. We're unsure. But Mitch, always a pleasure, my friend. And until next time, we'll see you on the next one.
- [08:15]
B
This podcast was produced in partnership with Amaze Media Labs.